American Journal of Gastroenterology最新文献

Objective: To perform a proof-of-concept study to evaluate the performance of individual and combined anorectal manometry (ARM) and balloon expulsion testing (BET) parameters for diagnosing dyssynergic defecation (DD) using machine learning (ML).

Methods: A single gastrointestinal motility specialist assessed (2008-2018) 307 patients. A machine learning model was constructed on ARM/BET data. BET was weight-based in 235 patients and time-based in 72 patients. Missing data were imputed. Data were split 75% train/25% test with preprocessing performed. Five common ML models were trained, and performance was compared using receiver-operating characteristic (ROC) area under the curve (AUC) and accuracy based on 100 bootstrapped samples. The most accurate models were fine-tuned and evaluated on the test set.

Results: The median age was 40 years, BMI 22.2 kg/m2, and 81% were female. Two-hundred and twenty-two (72.3%) patients had a clinical diagnosis of DD. Compared to patients without clinical diagnosis of DD, patients with DD had significantly lower rectoanal pressure gradient (RAG) (median -34.7 [IQR -57.1, -13.0] vs. -24.6 [IQR -44.5, -2.4] mmHg, p=0.009) and more abnormal BET (59.5% vs 11.8%, p<0.001). Four features were retained in the optimized model as predictors of DD: abnormal BET, greater resting anal pressure (RAP), and more negative RAG. The optimal parameter was abnormal BET (p=0.003), but combinations of two manometry results (RAP and RAG) yielded positive predictive values (PPV) >80%. The optimized logistic regression (LR) model had an AUC of 0.878, with a sensitivity of 75.0% and specificity of 81.8% at a probability threshold of 0.704.

Conclusion: Abnormal BET, or combination of two manometric parameters (RAP and RAG), provide >80% PPV for diagnosing DD in patients with chronic constipation.

Introduction: Flexible endoscopy (FE), rigid endoscopy (RE), and open surgery (OS) are established treatments for Zenker's diverticulum (ZD); however, comparative data on their clinical outcomes remain limited and inconsistent. We conducted the first multi-center matched cohort study comparing several procedural outcomes across FE, RE, and OS for ZD repair.

Methods: The study population was matched for age, gender, ASA score, and diverticulum. size resulting in Outcomes assessed included dysphagia scores, clinical and technical success, adverse events (AEs), procedure length, hospital length of stay (LOS), and need for further interventions. Both overall and pairwise comparisons were conducted. Subgroup analyses were performed to assess for comparative efficacy of FE procedure types, effect of prior intervention on procedure success, and the effect the year the procedure was performed on clinical success.

Results: There were 202 eligible patients over the study period. Propensity score matching resulted in 125 patients (FE n=44, RE n=36, OS n=45). FE demonstrated higher clinical success than RE (84.6% vs 55.6%, p=0.032) and similar success to OS (88.9%, p=0.712); OS showed higher success than RE (p=0.014). AE rates were lowest with FE (4.5%) compared to RE (8.3%) and OS (11.1%) though this was not statistically significant [p=0.212]. FE had shorter median LOS (0 days vs. 1 day for RE and 3 days for OS, p<0.001) and median procedure lengths (55 min vs. 58 min for RE and 78 min for OS, p=0.018).

Discussion: In this matched cohort, FE has clinical success comparable to OS and superior to RE for ZD repair, with shorter hospital stays, reduced procedure lengths, and a trend toward fewer adverse events. Our matched, multi-center design provides stronger comparative evidence than prior unmatched studies. These findings support FE as a first-line treatment for ZD repair, though prospective studies could further clarify long-term outcomes.

Introduction: Underwater endoscopic mucosal resection (UEMR) does not involve submucosal injection. Its ability to achieve sufficient resection depth comparable to that of conventional endoscopic mucosal resection (CEMR) remains unclear. This randomized controlled trial (RCT) aimed to compare UEMR and CEMR, with a specific focus on resection depth.

Methods: In this noninferiority RCT at four Japanese institutions, patients with colorectal tumors measuring 10-20 mm were randomly assigned to the UEMR or CEMR group. The primary outcome was resection depth analyzed per protocol, with a noninferiority margin of 400 µm.

Results: In total, 217 patients were randomly assigned to the UEMR (n = 109) or CEMR (n = 108) group. The mean resection depths were 1,688.9 ± 1,510.0 µm for UEMR and 1,432.3 ± 1,294.2 µm for CEMR, with a mean difference of 256.6 µm (95% confidence interval: -119.9 to 633.1 µm; P = 0.18). Subgroup analyses revealed that UEMR achieved significantly greater resection depths in the cecum and ascending colon (1,822.4 ± 1,469.7 µm vs. 1,096.5 ± 1,007.6 µm; P = 0.01) and in lesions with superficial morphology (1,238.7 ± 862.2 µm vs. 731.6 ± 376.1 µm; P ˂ 0.01) than CEMR. Nonexpert operators also achieved deeper resections with UEMR than with CEMR (1,786.6 ± 1,471.6 µm vs. 1,192.4 ± 1,074.7 µm; P = 0.07).

Discussion: UEMR achieved a resection depth comparable to that of CEMR. Given its consistent negative horizontal margins, UEMR may become the gold standard for resecting intermediate-sized colorectal tumors.